Reference-Offset Governor approach for the supervision of constrained networked dynamical systems

In this paper we propose a novel discrete-time predictive control strategy for the supervision of networked dynamical systems subject to coordination constraints. The approach joins in a unique framework Reference Governor (RG) and Offset Governor (OG) actions. RG units are used to modify the reference to a regulated plant whenever such an application would lead to constraint violations. On the contrary, OG units are used to add, for a finite amount of time, an offset to the nominal command at the input terminal in order to enlarge the set of feasible transients of the system. Numerical comparisons between the basic Reference Governor (RG) and the proposed Reference-Offset Governor (ROG) are presented. As a final example, we consider a well-known benchmark in power system control problems: a two-area power system subject to operative constraints on maximum deviations of frequencies, exchanged and generated powers.